Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A computer-implemented method comprising: determining, by a computing system, at least one salient point of interest in a frame of a content item based at least in part on a saliency prediction model, the saliency prediction model being trained to identify salient points of interest that appear in content items; generating, by the computing system, a barrel projection representation for the frame, wherein the barrel projection representation comprises portions including a circular top portion, a circular bottom portion, and a middle portion, wherein each of the portions includes at least one pixel of the frame; and applying, by the computing system, a view-based projection to the barrel projection representation for the frame, wherein the view-based projection enhances a quality in which a region corresponding to the at least one salient point of interest is presented.
This invention relates to computer-implemented methods for enhancing the presentation of salient regions in digital content, such as images or video frames. The problem addressed is the need to improve the visual quality of key areas in content, particularly when displayed on devices with limited viewing angles or non-linear projection surfaces. The method involves analyzing a frame to identify salient points of interest using a pre-trained saliency prediction model, which has been trained to detect visually important regions in content. Once these points are identified, the frame is transformed into a barrel projection representation, which divides the frame into three distinct portions: a circular top portion, a circular bottom portion, and a middle portion, each containing pixels from the original frame. This barrel projection is then processed using a view-based projection technique that enhances the visual quality of the regions corresponding to the salient points of interest. The goal is to ensure that these important areas are displayed with higher clarity and detail, improving the overall viewing experience. The method is particularly useful for applications where content is displayed on curved or non-flat surfaces, such as virtual reality headsets or wide-angle displays.
2. The computer-implemented method of claim 1 , wherein the plurality of distinct portions comprises a top portion of the frame associated with a first circular projection, a bottom portion of the frame associated with a second circular projection, and a middle portion of the frame associated with a rectangular projection.
This invention relates to a computer-implemented method for designing or analyzing a frame structure with distinct geometric projections. The method addresses the need for optimized frame designs that balance structural integrity, material efficiency, and functional adaptability. The frame includes multiple distinct portions, each featuring unique geometric projections to enhance performance. The top portion of the frame incorporates a first circular projection, which may improve load distribution or connection compatibility. The bottom portion includes a second circular projection, potentially serving similar or complementary purposes. The middle portion of the frame is equipped with a rectangular projection, which may provide additional structural reinforcement or facilitate integration with other components. These projections are strategically placed to optimize the frame's mechanical properties, such as stiffness, strength, or weight distribution. The method may involve simulating or analyzing the frame's behavior under various conditions to ensure it meets performance requirements. This design approach is particularly useful in applications where frames must support dynamic loads, such as in automotive, aerospace, or industrial equipment. The distinct projections allow for modular customization, enabling the frame to be adapted for different use cases while maintaining structural integrity.
3. The computer-implemented method of claim 2 , wherein the middle portion of the frame represents a middle 90 degrees field-of-view of a scene represented in the frame.
This invention relates to computer-implemented methods for processing visual data, specifically for capturing and representing a scene with a wide field of view. The problem addressed is the need to efficiently encode and transmit panoramic or wide-angle images while maintaining visual quality and reducing data redundancy. The method involves dividing a frame into distinct portions, where the middle portion corresponds to a 90-degree field-of-view of the scene. This central section is processed separately from other portions to optimize data handling, such as compression or transmission. The method may also include adjusting the resolution or encoding parameters of the middle portion to prioritize detail in the most critical viewing area. Additional processing steps may involve stitching multiple frames or applying distortion correction to ensure seamless integration of the wide-angle view. The invention aims to improve efficiency in capturing and transmitting high-resolution panoramic images, particularly for applications like virtual reality, surveillance, or immersive media.
4. The computer-implemented method of claim 2 , wherein applying the view-based projection to the barrel projection representation of the frame further comprises: increasing, by the computing system, a pixel density associated with a region in the middle portion of the frame that corresponds to the at least one salient point of interest.
This invention relates to image processing techniques for enhancing visual data, particularly in the context of projecting and analyzing images captured from curved or barrel-distorted perspectives. The problem addressed involves the need to improve the clarity and detail of regions of interest within such images, especially when those regions are centrally located. The method involves applying a view-based projection to a barrel projection representation of a frame. This projection process includes increasing the pixel density in a middle portion of the frame, specifically in regions corresponding to at least one salient point of interest. By enhancing pixel density in these areas, the method ensures that critical details within the central region of the image are preserved or improved, compensating for distortions introduced by the original barrel projection. This technique is particularly useful in applications such as surveillance, medical imaging, or any scenario where maintaining high fidelity in central image regions is essential. The approach dynamically adjusts pixel density to prioritize regions of interest, improving overall image quality and interpretability.
5. The computer-implemented method of claim 4 , wherein the region is stretched horizontally by a threshold amount.
This invention relates to image processing techniques for adjusting the aspect ratio of a region within an image. The problem addressed is the need to modify the dimensions of a specific area in an image while maintaining visual coherence, particularly when the original aspect ratio is not suitable for a target display or application. The method involves selecting a region within an image and stretching it horizontally by a predefined threshold amount. This adjustment ensures that the region fits within a desired aspect ratio without excessive distortion, preserving the overall visual quality. The stretching process may involve interpolation or other smoothing techniques to minimize artifacts. The method can be applied to various types of images, including photographs, graphics, or video frames, and may be used in applications such as digital photo editing, video processing, or display adaptation. The threshold amount for horizontal stretching is determined based on the original dimensions of the region and the target aspect ratio, ensuring a balanced and visually pleasing result. The technique may also include additional steps to refine the edges or boundaries of the stretched region to maintain sharpness and clarity. This approach provides a flexible solution for aspect ratio correction while avoiding the need for manual resizing or cropping.
6. The computer-implemented method of claim 2 , wherein applying the view-based projection to the barrel projection representation of the frame further comprises: decreasing, by the computing system, respective pixel densities associated with one or more other regions in the middle portion of the frame.
This invention relates to image processing techniques for optimizing the representation of panoramic or wide-angle images, particularly in applications like virtual reality (VR) or augmented reality (AR). The problem addressed is the inefficient use of pixel density in panoramic images, where regions in the middle portion of the frame often contain less critical visual information compared to peripheral regions. This leads to suboptimal storage and processing efficiency. The method involves applying a view-based projection to a barrel projection representation of a frame. The barrel projection is a type of wide-angle image projection that distorts the image to fit a wide field of view. The key improvement is dynamically adjusting pixel densities within the frame. Specifically, the method reduces pixel densities in one or more regions of the middle portion of the frame, where visual detail is less critical. This redistribution of pixel density allows for more efficient storage and processing while maintaining visual quality in areas of higher importance, such as the edges or peripheral regions. The technique can be applied in real-time systems, such as VR headsets, to optimize performance without sacrificing user experience. The method may also include additional steps like pre-processing the frame to identify regions of interest or post-processing to ensure smooth transitions between regions of varying pixel density. The overall goal is to enhance computational efficiency while preserving the visual integrity of the panoramic image.
7. The computer-implemented method of claim 6 , wherein the quality is enhanced in a portion of the plurality of distinct portions in the region.
This invention relates to computer-implemented methods for enhancing the quality of data or signals, particularly in systems where data is divided into distinct portions within a defined region. The problem addressed is the need to selectively improve quality in specific portions of the data or signal, rather than uniformly across the entire region, to optimize performance, reduce computational overhead, or focus on critical areas. The method involves processing a plurality of distinct portions within a region, where each portion represents a segment of the data or signal. The enhancement process is applied selectively to one or more of these portions, rather than uniformly across all portions. This selective enhancement allows for targeted improvements, such as increasing resolution, reducing noise, or correcting errors, in only the portions where it is most needed or beneficial. The method may involve analyzing the portions to determine which ones require enhancement, applying enhancement techniques like filtering, interpolation, or error correction to the selected portions, and then integrating the enhanced portions back into the overall region. By focusing enhancement efforts on specific portions, the method improves efficiency and effectiveness compared to blanket enhancements, which may waste resources on areas that do not require improvement. This approach is particularly useful in applications like image processing, signal processing, or data compression, where selective quality improvements can significantly enhance performance without unnecessary computational cost.
8. A system comprising: at least one processor; and a memory storing instructions that, when executed by the at least one processor, cause the system to perform: determining at least one salient point of interest in a frame of a content item based at least in part on a saliency prediction model, the saliency prediction model being trained to identify salient points of interest that appear in content items; generating a barrel projection representation for the frame, wherein the barrel projection representation comprises portions including a circular top portion, a circular bottom portion, and a middle portion, wherein each of the portions includes at least one pixel of the frame; and applying a view-based projection to the barrel projection representation for the frame, wherein the view-based projection enhances a quality in which a region corresponding to the at least one salient point of interest is presented.
This invention relates to a system for enhancing the presentation of salient points of interest in visual content, such as images or video frames. The system addresses the challenge of effectively highlighting important regions within a frame while maintaining visual quality. The system uses a saliency prediction model to identify key points of interest in a frame, which are areas likely to attract viewer attention. The model is trained to recognize patterns and features that make certain regions more salient than others. The system generates a barrel projection representation of the frame, which divides the frame into distinct portions: a circular top portion, a circular bottom portion, and a middle portion. Each portion includes pixels from the original frame. This projection helps in reorganizing the frame structure to better emphasize the salient regions. A view-based projection is then applied to the barrel representation, enhancing the presentation quality of the regions corresponding to the identified salient points. This projection adjusts the visual display to ensure that the important areas are more prominently and clearly visible to the viewer. The system dynamically processes frames to improve the overall viewing experience by focusing on the most relevant content.
9. The system of claim 8 , wherein the plurality of distinct portions comprises a top portion of the frame associated with a first circular projection, a bottom portion of the frame associated with a second circular projection, and a middle portion of the frame associated with a rectangular projection.
This invention relates to a system for securing a frame, such as a picture frame or a structural frame, using distinct projections to enhance stability and alignment. The system addresses the problem of maintaining proper positioning and preventing shifting or misalignment of frames, particularly in applications where frames are subjected to external forces or vibrations. The system includes a frame with multiple distinct portions, each designed to interact with corresponding projections. The top portion of the frame features a first circular projection, which provides a secure and rotatable connection point. The bottom portion of the frame includes a second circular projection, ensuring stability and preventing lateral movement. The middle portion of the frame incorporates a rectangular projection, which offers additional support and alignment. These projections work together to distribute forces evenly across the frame, reducing stress concentrations and improving overall structural integrity. The circular projections allow for rotational adjustment, while the rectangular projection ensures precise positioning and resistance to shear forces. This design is particularly useful in applications where frames must remain securely in place under dynamic conditions.
10. The system of claim 9 , wherein the middle portion of the frame represents a middle 90 degrees field-of-view of a scene represented in the frame.
This invention relates to a system for capturing and processing panoramic or wide-angle images, particularly focusing on the representation of a scene within a frame. The system includes a frame divided into distinct portions, where the middle portion specifically represents a 90-degree field-of-view of the scene. This middle portion is flanked by additional portions that may represent other segments of the scene, such as left and right side views, to provide a broader perspective. The system is designed to enhance the visualization of wide-angle or panoramic imagery by structuring the frame in a way that clearly delineates different angular segments of the scene. This approach may be useful in applications like surveillance, virtual reality, or augmented reality, where capturing and displaying a wide field of view is essential. The system ensures that the central 90-degree view is prominently displayed, allowing for detailed observation of the most critical portion of the scene while still providing context from adjacent angles. The invention may also include mechanisms for adjusting the field-of-view representation dynamically, depending on the application or user requirements.
11. The system of claim 9 , wherein applying the view-based projection to the barrel projection representation of the frame further causes the system to perform: increasing a pixel density associated with a region in the middle portion of the frame that corresponds to the at least one salient point of interest.
This invention relates to image processing systems for enhancing visual data, particularly in applications requiring detailed analysis of specific regions within a frame. The system addresses the challenge of maintaining high-resolution detail in areas of interest while efficiently processing the entire frame. The invention builds on a barrel projection representation of a frame, which distorts the image to reduce peripheral data while preserving central content. A view-based projection is applied to this representation, further refining the image by increasing pixel density in a middle portion of the frame corresponding to at least one salient point of interest. This selective enhancement ensures critical regions are captured with greater detail, improving accuracy in tasks like object recognition, tracking, or medical imaging. The system dynamically adjusts pixel density based on identified salient points, optimizing computational resources by focusing processing power where it is most needed. This approach is particularly useful in scenarios where high-resolution detail is required in specific areas, such as surveillance, autonomous navigation, or augmented reality, where peripheral data may be less critical. The invention improves upon traditional image processing methods by intelligently allocating resolution based on content relevance, reducing data redundancy and enhancing overall system efficiency.
12. The system of claim 11 , wherein the region is stretched horizontally by a threshold amount.
A system for image processing involves analyzing and modifying regions within an image to enhance visual quality or correct distortions. The system identifies a specific region within the image and applies a transformation to adjust its dimensions. In particular, the system stretches the identified region horizontally by a predefined threshold amount. This horizontal stretching can be used to correct aspect ratio distortions, improve readability of text, or adjust the layout of graphical elements. The stretching operation may be applied uniformly across the entire region or selectively to portions of the region based on additional criteria, such as edge detection or feature alignment. The system may also include preprocessing steps to analyze the image and determine optimal stretching parameters, ensuring the modification does not introduce unintended artifacts. The horizontal stretching can be combined with other transformations, such as vertical scaling or rotation, to achieve the desired visual effect. This approach is useful in applications like document scanning, medical imaging, or digital photography, where precise control over image dimensions is required. The system ensures that the stretching operation is applied in a controlled manner to maintain the integrity of the image content while achieving the intended correction or enhancement.
13. The system of claim 9 , wherein applying the view-based projection to the barrel projection representation of the frame further causes the system to perform: decreasing respective pixel densities associated with one or more other regions in the middle portion of the frame.
This invention relates to image processing systems for enhancing visual content, particularly in the context of barrel projection representations. The system addresses the challenge of optimizing image clarity and detail distribution in wide-angle or distorted images, such as those captured by fisheye lenses or panoramic cameras. The core functionality involves applying a view-based projection to a barrel projection representation of a frame, which adjusts pixel densities across different regions of the image. Specifically, the system reduces pixel densities in one or more regions within the middle portion of the frame. This adjustment helps balance visual quality by redistributing computational resources or visual emphasis, potentially improving sharpness or reducing artifacts in peripheral areas. The system may also include preprocessing steps to generate the barrel projection representation, ensuring compatibility with various input formats. The overall goal is to enhance image quality while maintaining computational efficiency, particularly for applications in surveillance, virtual reality, or high-resolution imaging.
14. The system of claim 11 , wherein the quality is enhanced in a portion of the plurality of distinct portions in the region.
The invention relates to a system for enhancing the quality of a region divided into multiple distinct portions. The system improves the quality of specific portions within the region, rather than uniformly enhancing the entire area. This selective enhancement allows for targeted improvements, optimizing resource usage and performance. The system may involve analyzing the region to identify portions requiring quality enhancement, applying specific processing techniques to those portions, and integrating the enhanced portions back into the overall region. The enhancement may involve adjusting resolution, clarity, signal strength, or other quality metrics depending on the application. The system is particularly useful in fields such as image processing, signal transmission, or data storage, where selective quality improvement is more efficient than uniform enhancement. By focusing on specific portions, the system avoids unnecessary processing of areas that do not require improvement, leading to better overall performance and resource efficiency. The invention ensures that only the necessary parts of the region are enhanced, providing a balanced and optimized solution.
15. A non-transitory computer-readable storage medium including instructions that, when executed by at least one processor of a computing system, cause the computing system to perform a method comprising: determining at least one salient point of interest in a frame of a content item based at least in part on a saliency prediction model, the saliency prediction model being trained to identify salient points of interest that appear in content items; generating a barrel projection representation for the frame, wherein the barrel projection representation comprises portions including a circular top portion, a circular bottom portion, and a middle portion, wherein each of the portions includes at least one pixel of the frame; and applying a view-based projection to the barrel projection representation for the frame, wherein the view-based projection enhances a quality in a portion of the plurality of distinct portions in which a region corresponding to the at least one salient point of interest is presented.
This invention relates to image processing techniques for enhancing the presentation of salient points of interest in content items, such as videos or images. The problem addressed is the need to dynamically adjust the visual representation of content to emphasize important regions, improving user engagement and perception. The method involves analyzing a frame of a content item to identify salient points of interest using a saliency prediction model. This model is trained to detect regions in content that are likely to attract attention. Once identified, the frame is transformed into a barrel projection representation, which divides the frame into distinct portions: a circular top portion, a circular bottom portion, and a middle portion. Each portion retains at least one pixel from the original frame. A view-based projection is then applied to the barrel projection representation. This projection selectively enhances the visual quality of the portion containing the salient point of interest, ensuring that the most relevant regions are presented with improved clarity or emphasis. The technique optimizes how content is displayed, particularly in applications where attention to specific details is critical, such as virtual reality, augmented reality, or adaptive video streaming. The approach leverages machine learning for saliency detection and geometric transformations to dynamically adjust the visual presentation.
16. The non-transitory computer-readable storage medium of claim 15 , wherein the plurality of distinct portions comprises a top portion of the frame associated with a first circular projection, a bottom portion of the frame associated with a second circular projection, and a middle portion of the frame associated with a rectangular projection.
This invention relates to a non-transitory computer-readable storage medium containing instructions for generating a frame structure with distinct portions. The frame includes a top portion with a first circular projection, a bottom portion with a second circular projection, and a middle portion with a rectangular projection. The frame is designed to support or interface with other components, likely in a mechanical or structural application. The distinct projections may serve functional purposes such as alignment, attachment, or structural reinforcement. The circular projections at the top and bottom may facilitate rotational or axial movement, while the rectangular projection in the middle could provide a flat surface for mounting or interfacing with other parts. The frame's design ensures stability and precise positioning of connected elements, addressing challenges in maintaining structural integrity and alignment in mechanical assemblies. The storage medium stores executable instructions to generate or fabricate this frame, enabling automated or digital manufacturing processes. The invention improves upon prior art by providing a more versatile and precisely defined frame structure with optimized projection shapes for specific applications.
17. The non-transitory computer-readable storage medium of claim 16 , wherein the middle portion of the frame represents a middle 90 degrees field-of-view of a scene represented in the frame.
This invention relates to image processing, specifically for capturing and representing wide-angle or panoramic views of a scene. The problem addressed is the efficient encoding and display of high-resolution panoramic images, particularly for virtual reality (VR) or augmented reality (AR) applications, where seamless and distortion-free viewing is critical. The invention involves a non-transitory computer-readable storage medium storing instructions that, when executed, perform a method for processing image frames. The method includes dividing a frame into multiple portions, where each portion corresponds to a specific field-of-view (FOV) of the scene. The middle portion of the frame represents a 90-degree FOV, providing a central, high-resolution view of the scene. Additional portions may represent other angular segments, such as 45-degree or 180-degree FOVs, depending on the application. The method further includes encoding these portions to optimize storage and transmission, ensuring that the panoramic image can be efficiently rendered in real-time for VR/AR devices. The encoding may involve adaptive compression techniques to balance quality and performance, particularly for dynamic scenes where real-time processing is essential. The system ensures that the middle portion, which is often the primary focus in VR/AR applications, is rendered with minimal distortion and high fidelity. This approach allows for efficient panoramic image handling while maintaining visual quality for immersive experiences.
18. The non-transitory computer-readable storage medium of claim 16 , wherein applying the view-based projection to the barrel projection representation of the frame further causes the computing system to perform: increasing a pixel density associated with a region in the middle portion of the frame that corresponds to the at least one salient point of interest.
This invention relates to image processing techniques for enhancing visual data, specifically improving the representation of salient regions in images or video frames. The problem addressed is the need to dynamically adjust pixel density in regions of interest to improve visibility and detail without uniformly increasing resolution across the entire image, which would be computationally inefficient. The system processes a barrel projection representation of a frame, which is a wide-angle or panoramic image format. A view-based projection is applied to this representation, focusing on at least one salient point of interest within the frame. The key innovation is increasing the pixel density in the middle portion of the frame that corresponds to the salient region, effectively zooming in or enhancing detail in that area while maintaining lower density in less critical regions. This selective enhancement improves computational efficiency and visual clarity for specific areas of interest, such as objects or features identified as important in the image. The method involves analyzing the frame to identify salient points, then dynamically adjusting pixel density in the corresponding region. This approach is particularly useful in applications like surveillance, virtual reality, or augmented reality, where certain regions require higher detail while others can be rendered at lower resolution. The technique optimizes both processing resources and visual quality by prioritizing regions of interest.
19. The non-transitory computer-readable storage medium of claim 16 , wherein applying the view-based projection to the barrel projection representation of the frame further causes the system to perform: decreasing respective pixel densities associated with one or more other regions in the middle portion of the frame.
This invention relates to image processing techniques for optimizing the representation of panoramic or wide-angle images, particularly in systems where efficient storage and display of such images are critical. The problem addressed is the distortion and data redundancy that occurs when capturing and processing wide-angle or panoramic images, which often results in uneven pixel density across the image, particularly in the central regions where distortion is most pronounced. The invention describes a method for processing image frames, where a barrel projection representation of a frame is generated to correct for wide-angle lens distortion. This representation is then transformed using a view-based projection to optimize the image for storage or display. The transformation includes reducing the pixel density in one or more regions of the middle portion of the frame, which helps minimize redundant data and improve processing efficiency. The system dynamically adjusts pixel density based on the content and distortion characteristics of the image, ensuring that visually important regions retain high fidelity while less critical areas are compressed or downsampled. The technique is particularly useful in applications such as virtual reality, augmented reality, and high-resolution imaging systems where maintaining image quality while reducing computational overhead is essential. By selectively adjusting pixel density, the method ensures that the final output is both visually accurate and efficiently processed.
20. The non-transitory computer-readable storage medium of claim 15 , wherein the applying the view-based projection to the barrel projection representation for the frame enhances the quality in the portion of the plurality of distinct portions by allocating increased number of pixels to the region.
This invention relates to image processing techniques for enhancing the quality of images captured by wide-angle or fisheye lenses, particularly in regions where distortion is most pronounced. The problem addressed is the inherent distortion and reduced resolution in peripheral regions of images captured with such lenses, which often results in poor image quality in those areas. The invention involves a method for processing image data to improve visual quality in specific regions of an image. A barrel projection representation of the image is generated, which maps the distorted image data into a format that preserves geometric relationships. A view-based projection is then applied to this representation, selectively enhancing the quality of certain portions of the image by allocating a greater number of pixels to those regions. This pixel allocation compensates for the distortion, effectively increasing resolution and clarity in the targeted areas. The method may also include additional steps such as correcting lens distortion, adjusting color balance, or applying sharpening filters to further refine the image. The technique is particularly useful in applications requiring high-quality wide-angle imaging, such as surveillance, virtual reality, and automotive cameras, where maintaining detail in peripheral regions is critical. By dynamically adjusting pixel distribution, the invention ensures that areas of interest retain high fidelity while minimizing artifacts introduced by the projection process.
Unknown
March 3, 2020
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.